Liren Large Software Subsidy 16

home *** CD-ROM | disk | FTP | other *** search

/ Liren Large Software Subsidy 16 / 16.iso / w / w020 / 1.ddi / INSTALL / README.TR1 < prev next >

Wrap

Text File | 1991-11-19 | 21.5 KB | 514 lines

README.TR1 ======= Preface ======= README.TR1 is an ASCII text file version of the TypeReader Supplementary Information documentation that came with TypeReader. README.TR1 contains information not included in either the TypeReader User's Guide or the TypeReader Scanner Installation Manual. README.TR1 is comprised of the following sections: "Updates to TypeReader User's Guide" describes changes and additions to the TypeReader User's Guide after the manual was printed. "TypeReader Notes" includes additional tips and techniques to help you when using TypeReader. The file README.TR2 contains the "Technical Supplement" section of the printed document. ================================== Updates to TypeReader User's Guide ================================== This section of TypeReader Supplementary Information describes the changes and additions to TypeReader and/or to the TypeReader User's Guide that were made after the manual was printed. It includes the following four changes: o New feature: "Disable Auto processing prompt" o Change in manual and usage note: Page displayed when Auto processing or Process Deferred Job is completed o Change in manual: Deferred Job Format o Usage note: Using Open... and Process Deferred Job New Feature: "Disable Auto processing prompt" ============================================= General description: "Disable Auto processing prompt" ----------------------------------------------------- "Disable Auto processing prompt" has been added to the User Preferences... dialog box. By default, after you've used Auto to scan one or more pages or to read in a file, TypeReader displays a dialog box that prompts you to continue in a variety of ways. You may choose to scan another page or stack of pages, scan the second side of a page or a stack, or read in another file. This lets you read in and process multiple files or stacks of pages as one document. However, it also means that you have to click End to proceed with Auto processing once Get Page is done. If instead you want to start the Auto process and have it proceed fully automatically, check "Disable Auto processing prompt" in User Preferences.... When you click Auto to start processing, TypeReader scans the page or stack in the scanner or reads in the first file you specify. Then, with no further intervention from you, Auto continues on with Locate and Recognize. Guidelines for using "Disable Auto processing prompt" ----------------------------------------------------- When you want to use Auto processing to read in more than one page, more than one stack of pages, or more than one file, make sure the "Disable Auto processing prompt" check box is not checked. When you want to use Auto processing to read in a single page, one stack of pages, or one file and continue automatically with Locate and Recognize, check the "Disable Auto processing prompt" check box. Steps for using Auto with and without "Disable Auto processing prompt" ------------------ When you use Auto with "Disable Auto processing prompt" off (not checked), processing occurs as described on pages 49 through 53 of the TypeReader User's Guide. Those steps are summarized here, with the new information inserted. pages 49-50: General steps to do Auto from a scanner 1. Place one or more pages in the ADF OR Place one page on the flatbed 2. Make sure "Disable Auto processing prompt" in User Preferences... is NOT checked 3. Click the Scanner icon in the Get Page section of the Gallery if it's not already selected 4. Check the Gallery to make sure that the rest of the Get Page settings, and the Locate and Recognize settings, are set the way you want them 5. Click Auto 6. If you're not done, go on to To scan the second side of double-sided pages or To scan additional pages OR If you're done, click End pages 52-53: General steps to do Auto from a file 1. Make sure "Disable Auto processing prompt" in User Preferences... is NOT checked 2. Click the File icon in the Get Page section of the Gallery 3. Check the Gallery to make sure that the Locate and Recognize settings are set the way you want them 4. Click Auto 5. Find and select the file you want 6. Click Get 7. To add another file on the end of the one you just read, repeat steps (5) and (6) as often as necessary OR When you're done reading files, click End When you use Auto with "Disable Auto processing prompt" on (checked), processing occurs as follows: page 53: To do Auto from a scanner without the Auto processing prompt 1. Place one or more pages in the ADF OR Place one page on the flatbed 2. Check "Disable Auto processing prompt" in User Preferences... 3. Click the Scanner icon in the Get Page section of the Gallery if it's not already selected 4. Check the Gallery to make sure that the rest of the Get Page settings, and the Locate and Recognize settings, are set the way you want them 5. Click Auto The entire TypeReader process of Get Page, Locate, and Recognize is completed automatically. The page(s) in the scanner are scanned and each page is located and recognized. When all pages are completed, the last page of the document is displayed in Image View at Actual Size or the last selected view size. page 53: To do Auto from a file without the Auto processing prompt 1. Check "Disable Auto processing prompt" in User Preferences... 2. Click the File icon in the Get Page section of the Gallery 3. Check the Gallery to make sure that the Locate and Recognize settings are set the way you want them 4. Click Auto 5. Find and select the file you want 6. Click Get The entire TypeReader process of Get Page, Locate, and Recognize is completed automatically. The pages in the specified file are read in and each page is located and recognized. When all pages are completed, the last page of the document is displayed in Text View at Actual Size or the last selected view size. Additional changes to the manual -------------------------------- On pages 147-148, in Save Settings As..., the state of the "Disable Auto processing prompt" check box is also saved. On pages 148-149, in Retrieve Settings..., the state of the "Disable Auto processing prompt" check box is also retrieved. On page 164, in User Preferences..., the state of the "Disable Auto processing prompt" is also set. In User Preferences..., the following new paragraphs are required: page 167: "Disable Auto processing prompt" setting By default, after you've used Auto to scan one or more pages or to read in a file, TypeReader displays a dialog box that prompts you to continue in a variety of ways. You may choose to scan another page or stack of pages, scan the second side of a page or stack, or read in another file. This lets you read in and process multiple files or stacks of pages as one document. However, it also means that you have to click End to proceed with Auto processing once Get Page is done. If instead you want to start the Auto process and have it proceed fully automatically, check "Disable Auto processing prompt" in User Preferences.... When you click Auto to start processing, TypeReader scans the page or stack in the scanner or reads in the first file you specify. Then, with no further intervention from you, Auto continues on with Locate and Recognize. Also on page 167, the following step is added: page 167: 6. Click in the checkbox to toggle "Disable Auto processing prompt" Also on page 167, the new User Preferences... dialog box, shown above, replaces the illustration in the manual. Change in manual and usage note: Page displayed when Auto processing or Process Deferred Job is completed ========================================================= General description of change ----------------------------- When Recognizing is completed, using either Auto processing or Process Deferred Job, the last [instead of first] page of the document is displayed at Actual Size or the last selected view size. To flip to the first page of the document, use Ctrl+Home. In the TypeReader User's Guide, the following pages are affected: page 54: When the document has been fully recognized, TypeReader switches to Text View. The last [instead of first] of the pages you just read is displayed at Actual Size or the last selected view size. To flip to the first page of the document, use Ctrl+Home. page 145: 5. Click Yes The last [instead of first] page of the document is displayed at Actual Size in Text View. Flip to the first page of the document by pressing Ctrl+Home. You can now preview the document, check it for suspects and illegibles, and edit it as needed. Change in manual: Deferred Job Format ===================================== On page 109, add the following paragraphs: Saving to Deferred Job Format Deferred Job Format is a special case of TypeReader (TIFF) format. Use it to save work in progress so the document may be reopened for further single-step processing (using Open...) or for final processing (using Process Deferred Job). When you save a document in Deferred Job Format, it is automatically saved in appropriate TypeReader (TIFF) format in the special DEFER directory. The current state of each page in the document is saved, including any Locate and Recognize information. See "Deferred Processing," "Create Deferred Job," and "Process Deferred Job" for more information. NOTE: When you want to save work in progress so that you can reopen it for further single-step processing, you may save in either Deferred Job Format or TypeReader (TIFF) format With Image and Text Regions (if there is no recognized text) or With Image and Formatted Text (if there is recognized text). The difference is that when you save a document in Deferred Job Format, it is automatically saved to the DEFER directory, where it is available to be processed using Process Deferred Job. Usage note: Using Open... and Process Deferred Job ================================================== There are two things you can do with a file saved in one of the TypeReader (TIFF) formats or in Deferred Job Format: o you can read it in for further single-step processing using Open... o you can complete automatic processing on it using Process Deferred Job NOTE: When you use Process Deferred Job, it looks for files only in the DEFER directory. Documents that are saved in Deferred Job Format are automatically saved to the DEFER directory. Documents saved in TypeReader (TIFF) formats may be saved to any directory. To use Process Deferred Job on a document saved in a TypeReader (TIFF) format, either save the file in the DEFER directory, or copy the saved file to the DEFER directory before you open it using Process Deferred Job. When you're reading a file in for further single-step processing, it's important to remember to use Open... in order to preserve any Locating or Recognizing you've already done. Do not use Get Page. When you use Open... to read in a file that you've saved in any TypeReader (TIFF) format or in Deferred Job Format, each page is retrieved with the Locate and Recognize information that was saved with it. See "Saving to Deferred Job Format" for more information. In contrast, if you use single-step Get Page to read in a file saved in any TypeReader (TIFF) format or in Deferred Job Format, the page is processed from scratch and any Locate and/or Recognize information in the document is discarded. CAUTION: Only use single-step Get Page, with source set to File, to read in files that have been saved in non-TypeReader TIFF and PCX formats. To read files that have been saved in any TypeReader (TIFF) format or in Deferred Job Format, use Open... or Process Deferred Job. ================ TypeReader notes ================ This section includes additional tips and techniques to help you when using TypeReader. Please refer to Chapters 12 and 13 of the TypeReader User's Guide for additional information on getting the most from TypeReader. Broken and touching characters ============================== TypeReader is good at recognizing characters that are broken or touching, especially when you use brightness level to compensate for bad character quality (see discussion in "Brightness control" in the TypeReader User's Guide). A broken character is one in which the structure of the character is degraded in some way. A common example of structural degradation is the incomplete loop that may be found in characters such as "p," "o," "q," and "b" when the characters are printed or copied too light. Touching characters often result from a combination of tight kerning and poor print quality of the source document. You can enhance TypeReader's ability to accurately recognize text that has broken or touching characters by manipulating the brightness level used during scanning. However, when you have both broken and touching characters on the same page or in the same document, a conflict may arise. A brightness setting that aids in recognizing touching characters will often aggravate the problem of broken characters. In such a situation, you'll usually find that it's better to decrease the brightness level (darken the image) to compensate for the broken characters, even though by doing so you may increase the frequency of touching characters. This works because TypeReader recognizes touching characters with a higher level of confidence than broken characters. Because broken characters are more of a problem than touching characters, you should decrease the brightness just enough to compensate for the broken characters. This will typically involve some experimentation, but you'll find that recognition usually can be dramatically improved by minimizing broken characters through brightness adjustments. High resolution images ====================== When using very high resolution images (>400 dots per inch) your computer may run out of memory, or even disk space. Additionally, TypeReader may have the following difficulties with very high resolution image files. Incorrect text regions ---------------------- At resolutions above 600 dpi, some TIFF and PCX files may have their text regions defined incorrectly when you Locate using the Normal setting or the Force Single Column setting. Incorrectly located or sequenced text regions can cause problems with recognition and the formatting of output. If text regions are not being Located correctly in high resolution text images, you can manually locate text regions in your document, instead of using Normal or Force Single Column. If you have many pages to process, you can define an appropriate template that you can then apply automatically. See "Locating, text regions, and templates" in the TypeReader User's Guide for more information on text regions. Inaccurate view sizes --------------------- Sometimes, when you view high resolution files at view sizes other than Actual Size, the size of the screen display may be five or ten percent larger than the image size selected with the View Size Selector. This has no effect on TypeReader's performance. Locating small, stylized fonts ============================== When you use Normal or Force Single Column to define text region rectangles on a page, TypeReader ignores any graphic elements and only creates text regions around text. When a page has text in a stylized 5 point font, TypeReader may incorrectly assume that this text is a graphic image. If you find that your small fonts have been overlooked during Locate, manually locate the additional text regions. (See "Locating text regions manually" in the TypeReader User's Guide.) After you've defined the text region(s), TypeReader should recognize the stylized 5 point font correctly as text. Check spelling vs. highlight word... ==================================== During Recognizing, TypeReader uses the built-in dictionary and the user dictionary you specify to enhance and verify recognition results. TypeReader checks every word to see if it's in either dictionary and keeps track of the words it doesn't find. You may choose to highlight these words in Text View in one of two ways--you can select the "Highlight word not in dictionary" setting in User Preferences..., or you can select Check Spelling from the Edit menu. Although they both highlight words they don't find in either dictionary, when you use Check Spelling, more words will be identified as not in the dictionaries than when you use "Highlight word not in dictionary." This happens because Check Spelling always treats punctuation marks as word delimiters. It therefore misinterprets hyphenated words at the ends of lines as two words and misidentifies terms that contain punctuation as ending before the punctuation mark. Often, these incorrectly interpreted character sequences will in fact not be in either dictionary and will be flagged by Check Spelling. "Highlight word not in dictionary" doesn't make these kinds of mistakes. It correctly interprets hyphenated words at the ends of lines and recognizes terms that contain punctuation marks. The following examples show two instances of text sequences that "Highlight word not in dictionary" correctly identifies and that Check Spelling misidentifies: and it doesn't recog- nize hyphenated words (1) sometimes longer than 25ft. and still growing (2) In example (1), "Highlight word not in dictionary" will know that "recog" and "nize" are two parts of the same hyphenated word and will find "recognize" in the built-in dictionary. Check Spelling will think that "recog" and "nize" are two words, neither of which it will find in either dictionary. In example (2), assume you've added the term "25ft." to your user dictionary. "Highlight word not in dictionary" will identify the term, including the period, and find it in the user dictionary. Check Spelling will discard the period and unsuccessfully look for "25ft" in both dictionaries. Mimimum system configuration issues =================================== The minimum amount of RAM that TypeReader will run in is four megabytes. Under normal circumstances, this amount of RAM is entirely adequate. However, there are several things you can do to guarantee the smooth operation of TypeReader with a 4mb system. TypeReader and Adobe Type Manager --------------------------------- Running Adobe Type Manager (ATM) on a system that is minimally configured (4mb RAM) in Windows Standard Mode can reduce TypeReader's recognition speed . This problem can be avoided by adding more RAM or by running Windows in 386 Enhanced Mode. You can gain additional recognition speed by limiting the amount of RAM that ATM uses for its font cache. The default cache size is 96K. We have found that ATM runs better with TypeReader when the font cache is reduced to 64k. You can reduce ATM's font cache by running the ATM Control Panel program (ATMCNTRL.EXE) and setting the value for Font Cache to 64k. See the Adobe Type Manager User Guide for further instruction on setting the font cache. TypeReader and disk caching --------------------------- If you find that TypeReader returns an "Out of memory" message when you are scanning legal and/or landscape pages, you may be able to solve this problem by reducing the size of Windows' disk cache (SMARTDRV.SYS). It is not unusual for a 4mb system to be using a 1mb cache under Windows, which may be more caching than you really need. You can (and should) experiment to find the best cache size for your system. During our testing on 4mb systems, we typically ran TypeReader with a disk cache of 256k. For further instruction on changing the size of your disk cache, please refer to chapter 13 of the Microsoft Windows User Guide. Maximum job size ---------------- There is a limit the the total number of pages you can process in TypeReader when using a 4mb system. This limit is in the 300 page range. If you will regularly work with jobs of 300 pages or more, we suggest that you increase the amount of RAM in your system to 6mb or greater. Exporting special characters ============================ Appendix B of the TypeReader User's Guide lists the characters that TypeReader is capable of recognizing. Some of the characters on this list are "special" in that they come from the extended ASCII symbol set. With all export file formats other than Rich Text Format(RTF) and AMI Professional 1.2, we convert these "special" characters to text equivalents. The following table lists these special characters by name and the converted text equivalents. Character Name On Export -------------- --------- cent c pound p yen YEN copyright (C) register (R) one-half 1/2 one-quarter 1/4 plus or minus +- divide by / bullet . degree deg pi (paragraph symbol) PI section S: